Browsing by Author "Lagerspetz, Mattias"

This thesis examines the drivers behind the formation of housing prices in Tallinn, Estonia. The focus is, more specifically, on the relation between changes in demographics and housing prices. There will be significant changes in Tallinn’s demographic composition in the years to come, as the number of births in Tallinn have declined steeply from the 1990s on. Past studies have shown, that most significant increase in people’s demand for housing occurs in their 20s, and as such the sharp fall in the number of young people in Tallinn, could result in a decrease of demand for housing. First, a model suitable for examining demographics’ effect on housing is introduced. In the model demand for housing by age-group is first estimated from cross-sectional data. Next, the obtained estimates are combined with demographic time series data to obtain a demographic demand time series. Lastly, the relations between housing prices and demographic demand (as well as other macroeconomic variables) are examined. Results from a number of previous studies employing a similar model are also presented. In the empirical section of this thesis, a similar model is applied on data from Tallinn, spanning from 1997 to 2014. Firstly, the demand for housing by age-group is estimated from cross-sectional data obtained by combining estimates for the value of almost 170 000 Tallinn housing units with the inhabitants of the units, as reported in Estonian 2011 census data. Next, the demand estimates are combined with time-series of past demographic data on Tallinn, to construct a total demographic demand time series variable. Thirdly, the long-run relationships between housing prices, demographic demand and other macroeconomic variables are examined using cointegration analysis. Tallinn housing prices are modelled in vector error correction model framework, along with the obtained demographic demand, GDP per capita, as well as other macroeconomic time series. Housing prices’ long run elasticity estimates obtained are generally in line with results from previous studies. However, the results seem quite sensitive regarding model specification. Importantly, effects of increases in demographic demand on housing prices are found to be negative for some of the alternative model specifications. The forecasted changes in future demographic demand are also of not a very great magnitude, so the predictions for future housing prices for the models are affected relatively modestly by the forecasted changes in demographic demand.